Skip to content

Benchmark

Agentset Embedding Elo

Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Higher scores are better.

Models measured
15
Scores recorded
15
Sources
1
IQ weight
0.71

Score distribution

1,499.8
Mean
1,490
Median
63.9
Std dev
1,414
10th percentile
1,586
90th percentile
1,366
Min
1,3661,605
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.71 for embedding models, measured over 15 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard